Getting the Data Right


Aver is keenly focused on helping healthcare organizations use bundled payments for episodes of care as a key value-based contracting tool. Aver begins by using historic data to model the underlying fee-for-service activity in the context of episodes of care. While it’s true our payer clients can access their data at any time, we still operate in the fee-for-service world, and payers don’t have a “lens” through which they can identify value-based-care opportunities. Aver’s Episode Snapshot tool provides that lens.

In addition, we have also found that payers would benefit from a stronger organizational structure around their bundled payment program. Most payers that are exploring or operating bundled payment programs have large teams of employees from many different departments working on different aspects of the program. But each member of that team spends perhaps 5-10% of their time on bundles, and there is not one person whose entire job is bundles.

At Aver, our entire focus is bundles. Aver provides clients an elite team of bundled payment experts, engineers, and analysts whose sole focus is ensuring their bundled payment program succeeds in improving quality and reducing costs. And once those goals are achieved, we relentlessly build bundled payment programs to scale.


"Aver provides clients an elite team of bundled payment experts, engineers, and analysts whose sole focus is ensuring their bundled payment program succeeds in improving quality and reducing costs. "


But before we get there, we must begin with the end goal in mind, and that means ensuring Aver and our clients are operating from a shared understanding of bundled payments and of use cases clients are considering for their bundle analysis. Aver works with payers and providers to implement PROMETHEUS, BPCI/Medicare, and custom episode definitions. As a result, we’ve become experts on all types of episode methodologies and we begin by sharing our experiences regarding the nuances and challenges of any bundle definition.

For example, PROMETHEUS takes a population health approach to episodes of care, which means it works best when we run all 97 episode definitions together. PROMETHEUS episodes associate with each other, have complications of each other, and costs are often split between episodes. This may cause an episode to “look” different or have a higher or lower cost than internal payer analyses might show. By discussing the philosophies behind different episode definitions, we can ensure we’re all operating from the same playbook as we move forward with developing hypotheses to explore in the bundle opportunity analysis.

Another way we begin with the end in mind is by helping clients think how they will want to sort or display their data. For example, if payers want to sort data by employer group, we need to include this data from the start. If clients want to search their data by county, we need that county field included and populated from the beginning. We also have conversations about the minimum required fields Aver needs in order to complete our analysis, as well as what happens if the data is not 100% complete. The process for attributing episodes to providers and facilities is highly complex and varies by episode definition. If only 60% of a client’s National Provider Identifier fields are populated, we may be unable to attribute 40% of episodes to a provider, which would substantially affect our analysis.

Aver is fortunate to have really smart clients who have a good understanding of their data and tend to have a keen sense of the episodes of care they want to target and the providers who will likely be good partners for early implementation. We work with clients to further build out these hypotheses, adding our own expertise as we develop this framework for the bundle opportunity analysis.

With these hypotheses in hand, the next step is the heaviest lift in this process: obtaining the data. Aver typically ingests two years of data, including member eligibility files, provider files, and all types of claims - medical, imaging, lab, and prescription drug. Not all clients send all this data. For example, prescription drug claims typically aren’t necessary for the types of episodes payers are likely to target for bundled payments first. But as payers look to scale in the future, and begin to consider bundled payments for chronic conditions or behavioral/mental health bundles, prescription drug spending plays a much larger role in these types of episodes than in the typical surgical procedural episode. With chronic conditions, such as diabetes, prescription drug spending may account for up to 80% of the long-term costs of disease management. We’ve also found that providers are starting to ask very detailed questions, including about medical and prescription drug cost split, before accepting a contract with downside risk. By being informed by  this data from the beginning, Aver is able to include it in our bundle opportunity analysis and payers are able to present to interested provider partners.

Once we’ve thought through all these issues, Aver begins a multistep data and analytics quality assurance framework. This framework consists of multiple checkpoints, including three that typically occur before Aver conducts its bundle opportunity analysis. First, we request a data dictionary from the client to ensure that all fields required to perform the desired analyses are included and to validate data formats. Next, we assess the completeness of the raw data set provided by the client at the field level, yielding a report of summary statistics for each field, including total rows per file, counts of value, percent missingness, and more. Finally, our mapped data checkpoint assesses the relationship within the data and the validity of transformations.

After we’ve confirmed the quality and completeness of the data, Aver is ready to conduct the bundle opportunity analysis. This analysis uses the payer’s data to illustrate the potential episodes of care to target and providers to approach about bundled payments. Future posts will discuss this analysis in greater depth, but this is really where Aver can test a payer’s, as well as our own internal, hypotheses about the best place to start a bundled payment program and/or where to expand. Our analysis often confirms a payer’s initial hypotheses and always provides additional episodes or providers for a payer’s consideration as they begin to build or grow their program.

Aver’s Episode Snapshot tool for identifying and considering episodes of care for bundled payments is data-led and data-driven, making is absolutely critical that we get the data right. Our multi-step quality assurance framework ensures that each client’s bundle opportunity analysis is based on sound, accurate data. And Aver’s elite team of bundled payment experts help to ensure clients understand the nuances of different episode definitions and helps guide clients through the process of identifying episodes and providers for their program. Then, our team turns its entire focus on ensuring each client’s bundled payment program succeeds in improving quality, reducing costs, and taking the program to scale. Our next post will explore how Aver’s bundle opportunity analysis identifies episodes of care to use in starting or expanding a bundled payment program.


Customer Delivery Team
Aver, Inc


Customer Delivery Team
Aver, Inc

Want fresh insights on Value-Based Care delivered right to your Inbox? Become a member of the Aver Insights Community